Capstone and Reporting
Capstone and Reporting
Commands only. What each step does, why it is built this way, and the judgment behind it are in the book.
Infrastructure tier
8 labs
≈ 4.5–6 h
Pure Python · no Docker
Windows · macOS · Linux
Labs in this chapter
What you'll be able to do
- Run the whole book end-to-end, chaining the four attribution engines into a single evidence graph for one operator.
- Turn evidence into claims, each carrying a statement, a type, its provenance, and a falsifier.
- Set a defensible confidence on every claim by a fixed rule rather than by feel, and state a low-confidence finding honestly as low.
- Assemble a decision-ready report with a bottom line, findings traceable to source, a what-would-change section, and an attribution boundary.
- Grade a report on coverage, provenance, calibration, and the overclaim count, watch a careless report mislead without stating a single false thing, and state where the whole method stops.
CHAPTER 15
Capstone and Reporting
The capstone chains every engine you have built over a single operator into an evidence graph, then produces a decision-ready report where each claim carries a statement, a source, a confidence, and the thing that would prove it wrong.
Written up in the book, commands and all.